Implicit US-ASCII encoding somewhere

[chrysn]

(Proposed resolution at the end, full story for reference)

On my system I get errors like this:

2022-02-03 16:14:01 +0100 Unexpected error while processing request: invalid byte sequence in US-ASCII
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-livereload-0.3.17/lib/rack/livereload/body_processor.rb:78:in `gsub!'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-livereload-0.3.17/lib/rack/livereload/body_processor.rb:78:in `block in process!'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-livereload-0.3.17/lib/rack/livereload/body_processor.rb:76:in `each'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-livereload-0.3.17/lib/rack/livereload/body_processor.rb:76:in `process!'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-livereload-0.3.17/lib/rack/livereload.rb:28:in `_call'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-livereload-0.3.17/lib/rack/livereload.rb:14:in `call'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/connection.rb:86:in `block in pre_process'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/connection.rb:84:in `catch'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/connection.rb:84:in `pre_process'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/connection.rb:53:in `process'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/connection.rb:39:in `receive_data'
        /home/chrysn/.gem/ruby/2.7.0/gems/eventmachine-1.2.7/lib/eventmachine.rb:195:in `run_machine'
        /home/chrysn/.gem/ruby/2.7.0/gems/eventmachine-1.2.7/lib/eventmachine.rb:195:in `run'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/backends/base.rb:75:in `start'
        /home/chrysn/.gem/ruby/2.7.0/gems/thin-1.8.1/lib/thin/server.rb:162:in `start'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-2.2.3/lib/rack/handler/thin.rb:22:in `run'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-2.2.3/lib/rack/server.rb:327:in `start'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-2.2.3/lib/rack/server.rb:168:in `start'
        /home/chrysn/.gem/ruby/2.7.0/gems/rack-2.2.3/bin/rackup:5:in `<top (required)>'
        /home/chrysn/.gem/ruby/2.7.0/bin/rackup:25:in `load'
        /home/chrysn/.gem/ruby/2.7.0/bin/rackup:25:in `<main>'

and plain Internal Server Error results in the browser.

As a workaround, I start with RUBYOPT="-KU -E utf-8:utf-8" kdwatch or LC_ALL="en_GB.UTF-8" kdwatch or LC_ALL="C.UTF8: kdwatch, which makes things work better. (Setting LC_ALL to en_US.UTF-8 was insufficient, probably because that locale is not built on my system).

I think that my work environment is all set up for UTF-8 (which, really, all system should have been for ages):

$ env |grep LC\\\|LANG
LANG=en_GB.UTF-8
LANGUAGE=en_GB:en

Operating system is Debian GNU/Linux sid, in a setup that only builds the en_GB.UTF-8 (and a few selected other) locales.

---

I think that this is related to the ENV["LANG"]="en_US.utf-8" workaround on top of bin/kdwatch -- this sets the default to some (on my system), which I can override by setting LC_ALL.

Quite possibly, the workaround has become obsolete anyway; if it has not, C.UTF-8 might be a suitable replacement (as it should be available no matter which locales are built).

[cabo]

It seems C.UTF-8 is a recent invention. It is certainly not universal:

https://unix.stackexchange.com/questions/597962/how-widespread-is-the-c-utf-8-locale

What is the best way to check whether a locale is implemented? I could cycle through a few en_XX.UTF-8 at the start.